Torroid Core and Hall Effect device.

[dr pepper]

Oh my head hurts.
Ok that makes sense.
I dont remember doing anythign like that before.
If I need to calc a dc choke I design the choke as if it was for ac, after all the choke only does something when theres a change in voltage.

This tells you the flux density within a dc choke, its just a jiggled around equation, some textbooks will bend your head.
Wb = V * (time taken to reach max current) / Turns

A shunt might be better to detect the steering motor stall.
You can detect a sudden short with a choke, I remember a flashover detector from a rf welder, it just detected voltage across a choke, if the voltage went over a threshold it tripped, only a sudden change such as that from a flashover would trigger it, a steering motor isnt going to give a sharp edge though, a shunt and op amp would be easier to develop.

 

[MrAl]

Oh my head hurts.
Ok that makes sense.
I dont remember doing anythign like that before.
If I need to calc a dc choke I design the choke as if it was for ac, after all the choke only does something when theres a change in voltage.

This tells you the flux density within a dc choke, its just a jiggled around equation, some textbooks will bend your head.
Wb = V * (time taken to reach max current) / Turns

A shunt might be better to detect the steering motor stall.
You can detect a sudden short with a choke, I remember a flashover detector from a rf welder, it just detected voltage across a choke, if the voltage went over a threshold it tripped, only a sudden change such as that from a flashover would trigger it, a steering motor isnt going to give a sharp edge though, a shunt and op amp would be easier to develop.

Hi,

Yes the total flux is the ac component plus the dc component, and the dc component is often a pain because an ac choke is not made to have much dc or else it saturates. Some can saturate with just a small amount of dc unless they are designed for such. Power supply switchers almost always have to have a choke that can stand a significant dc current because they have to, after all, put out dc current, which means there is always dc current in the choke.
Transformers for example are usually designed for ac only, so the presence of only a small dc current can saturate the primary can cause all sorts of problems including blown transistors. For partial saturation, often the transformer makes a horrible noise because the flux is going too high for that core. You dont want to stay in the same room with a 10kwatt converter with a partially saturating core for very long

 

[dr pepper]

Your quite correct.

I treat chokes as ac even when used on dc hence the equation, I've got away with it so far.

Never noticed borderline saturation causing noise, whats going on there then?, some generated harmonic moving the windings or soemthing.

Pc supplies and a lot of switchers use iron core chokes for dc smoothing, iron powder has a smooth saturation curve, whereas power ferrite has a much sharper knee.
I put together a saturation testing rig for chokes, essential if like me you salvage stuff, one day I'll get around to putting a B/H curve tracer together.

 

[MrAl]

Hi,

Yes vibration when the waveform starts to clip, as that creates higher harmonics than is usually present. This happens mostly in line operated stuff.

 

[dr pepper]

Yes of course lower harmonics.
Never had that before, but I have had current draw modulation creating an audio envelope generating annoying noises, doesnt happen as much with current mode.
I used a tny series single chip smps for a project, it made a racket, turned out the current draw being low caused the chip to be on the threshold of startup and run mode, causing a series of clicks sounding like a weedwacker.

 

[MrAl]

Hi,

Oh that's funny

We had the modulation issue in several converters too, where the DC current in the primary would ramp up and down and thus the noise would come and go, at intervals of maybe 2 seconds ro something. The solution was to integrate the primary voltage and thus average over some cycles, then feed that back to the control circuit which would then even out the transistor 'on' times to decrease any DC offset in the primary. The laminations would vibrate and so did the windings. After varnishing the noise would die down somewhat, but still too loud so it had to be addressed eventually with the new circuit.

I havent actually worked first hand with the tny chips, most of the stuff i ever did was with a full transistor H bridge.

 

[dr pepper]

Playing with the compensation network can do a similar job.
One thing the tny chip series does is introduce within restraints a certain randomness to the switching freq, this is sposed to reduce audible noise and moreover switching emi.

 

[MrAl]

Hi,

That sounds like a good idea, if the output is DC. Might work with the AC converters too though.
The integration network was an addition to the compensation circuit, so it was really a special kind of compensation network that measured the primary rather than the secondary of the transformer. The normal feedback came from the secondary as that was what had to be regulated. The transformer action however blocks DC so the error offset never appears at the secondary, just at the primary, and that offset is what causes the problem.
We also incorporated transistor current detection later on. That detected a fault in the transistor current in both sides of the bridge. Some of those transistors were 100 bucks each

 

[kinarfi]

This is a current sensor salvaged from my spa with it's LM7808 power supply for a hall effect device that was replaced with a Honeywell SS495s device
My next thing to do is to see what type of air coil I need to make so my power steering unit can operate and yet be tripped of line when the current surges as it hits the end of travel, this has to happen for left and right turns, so the sensor has be prior the switching mechanism or 2 sensors placed back to back.
Jeff